Researchers have long been attempting to exploit the ability of targeting moieties or ligands to bind to specific cells (via receptors or otherwise) to target compositions such as detectable labels or therapeutic agents to particular tissues of an animal (especially a human). In such situations, the ability of the targeting moiety to bind to the target (e.g., affinity, avidity, and/or specificity) significantly impacts the ability to successfully target the desired tissues.
Numerous attempts have been made to use natural (e.g. polyclonal) and monoclonal antibodies, as targeting moieties in vivo. However, use of such antibodies present certain disadvantages, such as unacceptable levels of antigenicity—even for humanized antibodies. In addition, natural antibodies are difficult to produce in recombinant form, due to the number of chains, disulfide bonds, and glycosylation. Natural antibodies also present pharmacokinetic problems. Antibodies pose significant problems in imaging and radiotherapeutic applications because, due to their large size, accumulation in extravascular target tissue and clearance from the vascular system are both slow. This problem is especially critical when dealing with solid tumors, which present additional barriers to the ingress of large blood born compounds. Similar problems occur with antibodies used for imaging using other modalities, such as magnetic resonance imaging (MRI), ultrasound and light. If the antibody is radiolabeled with a diagnostic or therapeutic radionuclide, lower target to background ratios result in the images. In addition, an undesirable distribution of radiation exposure between the tumor and normal tissues occurs.
In attempts to solve these problems, efforts have been directed towards the construction of smaller entities with similar binding affinities using the essential features of the natural antibody binding regions. The building blocks are typically single-chain Fv fragments (scFv) which are monovalent. Combining fragments of this type so that they have the bivalent or multivalent properties of the antibodies has been problematic. In order to dock to a surface it is an advantage that the two binding sites on the antibody are connected via a flexible hinge to the constant region. Thus, in order to imitate the binding efficacy of antibodies, not only must the binding site be recreated, but so also must the bivalency (or higher valency) and the flexibility. This flexibility is needed because the protein backbone that makes up the nonbinding region of the scFv is still bulky relative to the binding site. Once an appropriate method has been devised to join two scFv fragments together, different scFv fragments can be joined together as well as more than the customary two scFv moieties present in natural antibodies. Certain scFv fragments, depending both on the VH/VL interface and the linker length, can spontaneously dimerize or multimerize. These “diabodies” are smaller than the natural antibody and do not have the immunological properties of the Fc portion (which activates complement and/or binds to Fc receptors), which they lack. The two (or more) binding sites are rotated relative to each other, and thus the antigen must be correctly positioned to accommodate this presentation.
“Miniantibodies” have properties similar to those of diabodies, but rather than a short 5-20 amino acid linker they have a relatively more flexible linker that allows freer orientation of the binding sites relative to each other, similar to in a natural antibody. Like diabodies, miniantibodies do not have the high molecular weight, immunologically active Fc dimer fragment. They can also be made by bacterial systems. Although they have desired advantages over natural antibodies, miniantibodies still suffer from having a relatively large size, which affects their pharmacokinetics, and must be made using biological methods. The smallest miniantibody is about 120 kDa in size.
Attempts have been made to use bispecific antibodies (e.g. antibodies that bind to two separate targets) to overcome one of the major deficiencies of antibodies, namely, that the size of the antibodies slows accumulation in the extravascular target tissue and clearance from the blood. The bispecific approach taken has been referred to as “pretargeting.” This approach uses a two-step protocol. A bispecific antibody with at least one arm that recognizes a tumor-associated antigen and at least one other arm that recognizes an epitope on a diagnostic or therapy agent is given as a first injection. After the unbound antibody has substantially cleared non-target tissues and has reached a maximum level in the tumor, the smaller, bispecific antibody-recognizable diagnostic or therapeutic agent is given. It is hoped that the latter agents distribute more rapidly throughout the body, and either bind to the bispecific antibody localized at the tumor, or are cleared via the kidneys.
An alternative to this approach attempts to use a mixed antibody avidin/biotin system in a two-step procedure. For example, a targeting antibody is conjugated with either avidin or biotin and then is injected whereupon it localizes in the tumor of interest. Thereafter, either biotin or avidin (depending on which was coupled to the targeting antibody), bearing an imaging or radiotherapeutic radionuclide, is injected and becomes localized at the site of the primary antibody by binding to avidin or biotin respectively.
Another approach to the use of antibodies as targeting moieties for radiopharmaceuticals or other diagnostic imagining agents has attempted to use a bivalent hapten to increase the avidity for the cell bound bispecific antibody over that of the circulating antibody. This approach relies on bidentate binding occurring with the cell bound antibodies, because the surface density on the cells is sufficiently high, but not occurring with the circulating antibodies, because the concentration is too low. In effect, the system makes use of the increase in avidity caused by the closer presentation of the antibodies/antigen on the cells.
Peptides have also been used as targeting moieties. In an attempt to improve the binding bi-specific peptide constructs have been prepared with two or more peptide based targeting agents selective for different targets. For example, a hybrid peptide having ligands to two targets selected from the somatostatin-, GRP-, CCK-, Substance P-, or VIP receptor and αvβ3 integrin was reportedly made and tested for the ability to bind to tumor cells. The initial evaluation showed no improved tumor uptake for the multiple ligand systems investigated. The investigators assumed that steric impairment leads to a reduction of the receptor affinities of the dimeric structures. Others have tested an RGD-DTPA-Octreotate hybrid peptide targeted towards both the αvβ3 integrin and the somatostatin-2 receptor for the ability to increase the tumor uptake over that of a peptide selective for one or the other targets. The different binding affinities of the two targeting moieties towards their targets, blood vessels and tumor cells, respectively, resulted in the avidity for tumors being dominated by the stronger (somatostatin mediated) interaction.
A variation of these approaches uses a bispecific diabody targeted to two different epitopes on the same antigen. This approach attempts to increase the avidity of the construct for the target, because, although the binding is monovalent for each epitope, the construct as a whole is bivalent to its target, as each of the binding epitopes is located within the same target molecule. In the case of the single molecule target, scFv fragments have been found to have insufficient affinity and an increase in avidity was required.
Two rationales underlie the approaches described above. The first rationale uses two different targeting moieties to overcome some of the pharmacokinetic problems associated with the delivery of antibodies to solid tumors. The second rationale uses two different targeting moieties to increase the avidity of the construct for a given target, such as a single molecule or a whole tumor. However, all of the approaches described suffer from various drawbacks. Thus, there remains a need for diagnostic and therapeutic agents with increased affinity and or avidity for a target of interest. There also remains a need for diagnostic and therapeutic agents that, when administered in vivo to a mammal, have acceptable pharmacokinetic properties.
Angiogenesis, the formation of new blood vessels, occurs not only during embryonic development and normal tissue growth and repair, but is also involved in the female reproductive cycle, establishment and maintenance of pregnancy, and repair of wounds and fractures. In addition to angiogenesis that occurs in the normal individual, angiogenic events are involved in a number of pathological processes, notably tumor growth and metastasis, and other conditions in which blood vessel proliferation is increased, such as diabetic retinopathy, psoriasis and arthropathies. Angiogenesis is so important in the transition of a tumor from hyperplastic to neoplastic growth, that inhibition of angiogenesis has become an active cancer therapy research field.
Tumor-induced angiogenesis is thought to depend on the production of pro-angiogenic growth factors by the tumor cells, which overcome other forces that tend to keep existing vessels quiescent and stable. The best characterized of these pro-angiogenic agents is vascular endothelial growth factor (VEGF) (Cohen et al., FASEB J., 13: 9-22 (1999)). VEGF is produced naturally by a variety of cell types in response to hypoxia and some other stimuli. Many tumors also produce large amounts of VEGF, and/or induce nearby stromal cells to make VEGF (Fukumura et al., Cell, 94: 715-725 (1998)). VEGF, also referred to as VEGF-A, is synthesized as five different splice isoforms of 121, 145, 165, 189, and 206 amino acids. VEGF121 and VEGF165 are the main forms produced, particularly in tumors (see, Cohen et al. 1999, supra). VEGF121 lacks a basic domain encoded by exons 6 and 7 of the VEGF gene and does not bind to heparin or extracellular matrix, unlike VEGF165.
VEGF family members act primarily by binding to receptor tyrosine kinases. In general, receptor tyrosine kinases are glycoproteins having an extracellular domain capable of binding one or more specific growth factors, a transmembrane domain (usually an alpha helix), a juxtamembrane domain (where the receptor may be regulated, e.g., by phosphorylation), a tyrosine kinase domain (the catalytic component of the receptor), and a carboxy-terminal tail, which in many receptors is involved in recognition and binding of the substrates for the tyrosine kinase. There are three endothelial cell-specific receptor tyrosine kinases known to bind VEGF: VEGFR-1 (Flt-1), VEGFR-2 (KDR or Flk-1), and VEGFR-3 (Flt4). Flt-1 and KDR have been identified as the primary high affinity VEGF receptors. While Flt-1 has higher affinity for VEGF, KDR displays more abundant endothelial cell expression (Bikfalvi et al., J. Cell. Physiol., 149: 50-59 (1991)). Moreover, KDR is thought to dominate the angiogenic response and is therefore of greater therapeutic and diagnostic interest (see, Cohen et al. 1999, supra). Expression of KDR is highly upregulated in angiogenic vessels, especially in tumors that induce a strong angiogenic response (Veikkola et al., Cancer Res., 60: 203-212 (2000)). The critical role of KDR in angiogenesis is highlighted by the complete lack of vascular development in homozygous KDR knockout mouse embryos (Folkman et al., Cancer Medicine, 5th Edition (B. C. Decker Inc.; Ontario, Canada, 2000) pp. 132-152).
KDR (kinase domain region) is made up of 1336 amino acids in its mature form. The glycosylated form of KDR migrates on an SDS-PAGE gel with an apparent molecular weight of about 205 kDa. KDR contains seven immunoglobulin-like domains in its extracellular domain, of which the first three are the most important in VEGF binding (Cohen et al. 1999, supra). VEGF itself is a homodimer capable of binding to two KDR molecules simultaneously. The result is that two KDR molecules become dimerized upon binding and autophosphorylate, becoming much more active. The increased kinase activity in turn initiates a signaling pathway that mediates the KDR-specific biological effects of VEGF.
Thus, not only is the VEGF binding activity of KDR in vivo critical to angiogenesis, but the ability to detect KDR upregulation on endothelial cells or to detect VEGF/KDR binding complexes would be extremely beneficial in detecting or monitoring angiogenesis. Diagnostic applications, such as detecting malignant tumor growth, and therapeutic applications, such as targeting tumoricidal agents or angiogenesis inhibitors to the tumor site, would be particularly beneficial.
Hepatocyte growth factor (also known as scatter factor) is a multi-functional growth factor involved in various physiological processes such as embryogenesis, wound healing and angiogenesis. It has become apparent that HGF, through interactions with its high affinity receptor (cMet), is involved in tumor growth, invasion and metastasis. In fact, dysregulated cMet expression (for example, the overexpression of cMet in neoplastic epithelium of colorectal adenomas and in other carcinomas as compared to normal mucosa) and/or activity, as well as hyperactivity of the cMet receptor through an autocrine stimulatory loop with HGF, has been demonstrated in a variety of tumor tissues and induces oncogenic transformation of specific cell lines.
In general, HGF is produced by the stromal cells, which form part of many epithelial tumors; however, it is believed that the production of HGF by tumor cells themselves comprises the main pathway leading to the hyperproliferation of specific tumors. HGF/cMet autocrine stimulatory loops have been detected in gliomas, osteosarcomas, and mammary, prostate, breast, lung and other carcinomas.
Interrupting the HGF interaction with the cMet receptor slows tumor progression in animal models. In addition to stimulating proliferation of certain cancer cells through activation of cMet, HGF also protects against DNA-damaging agent-induced cytotoxicity in a variety of cell lines susceptible to hyperproliferative phenotypes (e.g., breast cancer). Therefore, preventing HGF from binding to cMet could predispose certain cancer cells to the cytotoxicity of certain drugs.
In addition to hyperproliferative disorders, cMet also has been linked to angiogenesis. For example, stimulation of cMet leads to the production of vascular endothelial growth factor (VEGF), which, in turn, stimulates angiogenesis. Additionally, stimulation of cMet also has been implicated in promoting wound healing.
In addition to identifying the cMet receptor as a therapeutic target for hyperproliferative disorders, angiogenesis and wound healing, the large discrepancy between expression levels of neoplastic and corresponding normal tissues indicates that cMet is an attractive target for imaging applications directed to hyperproliferative disorders.